Available online at www.sciencedirect.

com

ScienceDirect
Procedia CIRP 48 (2016) 134 139

23rd CIRP Conference on Life Cycle Engineering

Effect of process parameters on mechanical recycling of

glass fibre thermoset composites
Norshah Aizat Shuaib*, Paul Tarisai Mativenga
School of Mechanical, Aerospace and Civil Engineering, The University of Manchester, Manchester M13 9PL, United Kingdom

* Corresponding author. E-mail address: norshah.shuaib@manchester.ac.uk

Abstract

High demand of glass fibre reinforced thermoset composites has led to manufacturing and end of life waste. Mechanical recycling is currently a
mature, rapid process in recycling such waste at an industrial scale. Unlike manufacturing processes, the effect of key process variables on
recyclate quality is not well understood. In this study analysis of variance was used to establish the key mechanical granulator process variables
that influence energy demand and recyclate quality. Two different granulator technologies were also compared. This information is vital in
selecting conditions for running recycling processes and in assessing the potential market for the generated recyclate.

2016 The Authors. Published by Elsevier B.V.

2016 Published by Elsevier B.V This is an open access article under the CC BY-NC-ND license
Peer-review under responsibility of the scientific committee of the scientific committee of the 23rd CIRP Conference on Life Cycle
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Engineering.
Peer-review under responsibility of the scientific committee of the 23rd CIRP Conference on Life Cycle Engineering

Keywords: composites; recycling; mechanical recycling;

1. Introduction Taken together these recycling drivers and challenges imply

Composite materials are in demand in critical industrial
sectors such as aerospace, automotive and wind energy due to
their lightweight, excellent fatigue resistance and high
corrosion resistance properties [1]. Glass fibre reinforced
plastic (GFRP) composites constitute around 98% of
composites by volume in the United Kingdom and European
composites production [2, 3] and present a major legacy
challenge for composite waste management.
The heterogeneous nature of composite material makes
recycling very challenging. In addition, most composites used
in mechanically demanding applications use thermoset matrix
which cannot be melted and remoulded. Currently, the main
disposal route for composite waste in the UK is through
landfill [4, 5]. With increasing legislation pressure for a
landfill ban, advancement of composite recycling technologies
is needed. According to End of Life Vehicle Directive
(2000/53/EC) the allowance for landfilling of end of life
vehicles reduces from 15% in 2006 to only 5% in 2015 [6].
that large volume recycling of GFRP needs urgent attention.
Current recycling technologies for glass fibre composites
are divided into mechanical, thermal, chemical and electrical
methods. Mechanical recycling involves size reduction of
composite waste into different size of recyclates through
milling processes [7]. The most common technique is a
hammer mill process [8, 9]. The basic principle of the hammer
mill machine is that the material is downsized through impact
and shear action until the fragments can pass through
predefined size of milling screen holes. Generally, the
recyclates are in forms of flakes, fibre-rich and resin-rich
fractions. Thermal methods consist of fluidised bed and
pyrolysis processes. Conventional and microwave pyrolysis
disintegrate the matrix part of composite material in the
absence of oxygen using oven and electromagnetic radiation
heating respectively, at a temperature around 300-700 C [10,
11]. In a fluidised bed process, composite waste is heated
rapidly using an air stream to enable matrix decomposition
[12]. Chemical recycling immerses glass fibre composite
waste in a suitable solvent such as water, acid and alcohol at a

2212-8271 2016 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the 23rd CIRP Conference on Life Cycle Engineering
doi:10.1016/j.procir.2016.03.206
 Norshah Aizat Shuaib and Paul Tarisai Mativenga / Procedia CIRP 48 (2016) 134 139
particular pressure and temperature to liberate the fibres [13].
However, thermal and chemical methods are best suited for
recycling of glass fibre composites due to low temperature
resistance of the glass fibres, which leads to severe fibre
strength degradation [13, 14]. Rouholamin et al [15] used
electrical discharge to disintegrate composite waste in an
aqueous solution. In a study by Hohenstein Institute in
Germany, microorganisms were used to degrade matrix part of
a composite [16]. However, these processes have low
processing rate. To date, mechanical method is the industrial
scale process available for recycling GFRP waste [7].
Recyclates obtained from mechanical recycling can be
incorporated in close loop or cross sector applications. For
instance, Filon Products and Hambleside Danelaw recycle
their in-house manufacturing waste and incorporate it back to
valley gutter products [17]. It has also been reported that,
mechanical strength of polymer based mortar products can be
improved by incorporating around 4-8% by weight of glass
fibre recyclates to replace sand aggregate fillers [18]. Despite
these few promising applications, there is an urgent need to
widen the market opportunities of mechanically recycled
GFRP.
The hammer mill technique has been extensively used in
the agricultural sector for size reduction of biomass materials
such switchgrass, corn stover and wheat straw [19, 20].
Factors that influence the energy performance of hammer mill
machines are divided into machine and material related
parameters [19-21]. Examples of the machine parameters are
screen size and operating speed. Material parameters depend
on the throughput, initial and final product size. These
parameters affect internal circulation and residence time of the
material inside the cutting chamber.
In mechanical recycling of composites, effect of operating
parameters on process energy demand and recyclate quality
was a knowledge gap identified for this research. This
information is important to assess reusability of the recyclates
in potential close loop or cross sector applications.
1.2. Research aim
This studys aim was to investigate the effect of
operational parameters on process energy demand and quality
of recyclates in mechanical recycling of glass fibre
composites. Three control factors investigated were hammer
mill screen size, material thickness and material size.
Performance of two different granulator technologies was also
compared. The vision is to develop the knowledge base for
selecting optimum parameters to minimise energy footprint
and to predict recyclate quality.
2. Methodology
Mechanical recycling glass fibre reinforced unsaturated
polyester waste was done using a Wittmann MAS1
granulator. The idea was to investigate effect of operational
parameters based on Taguchi experimental design. Selected
responses were specific process energy demand and
associated recyclate characteristics. Electrical power and
material residence time in the cutting chamber were used as a
135
basis in determining the overall energy demand. The
recyclates from each trial were analysed in terms of resin
content, fibre length distribution and weight fraction based on
particle size. Two different mechanical recycling technologies
were compared by processing GFRP scrap waste from a boat
manufacturer.
3. Experimental Procedure
3.1. Effect of operational parameters
A Wittmann MAS1 granulator, as shown in Figure 1 was
used. The machine has a 180 mm rotor diameter and is
powered by a motor rated at 2.2 kW. The rotational speed was
200 rpm. The granulator screen was located below the cutting
chamber and was interchanged into sizes of 4 mm and 6 mm.
The gap between the granulator blades and the chamber was
about 5 mm.
In this study, operational parameters selected to be studied
were plate thickness, plate size and screen size. The
experimental design was an L4 Taguchi orthogonal array and
the runs are shown in Table 1. Each experimental run was
repeated three times. The material used in this study was glass
fibre reinforced unsaturated polyester panels, manufactured
from glass fibre mats with random fibre orientation. The
panels were manufactured by hand lay-up technique and
supplied by Production Glassfibre. Average thicknesses of the
panels were 3 mm and 5 mm. The panels were cut into two
different sizes using a diamond tile cutter.
Fig. 1. Wittmann MAS1 granulator (a) machine;
(b) rotor blades cutting chamber [22]
Table 1. L4 Taguchi orthogonal array
Run Plate thickness Screen size Plate size
T1 3 mm 4 mm 45mm x 40 mm
T2 3 mm 6 mm 45 mm x 60 mm
T3 5 mm 4 mm 45 mm x 60 mm
T4 5 mm 6 mm 45mm x 40 mm
Throughout all trials, the 3-phase voltage and current were
recorded using a Fluke 434 power meter. The meter was
clamped to the three phase wires of the granulator. The basic
(no-load) power was measured before the panel insertion in
every trial.
136 Norshah Aizat Shuaib and Paul Tarisai Mativenga / Procedia CIRP 48 (2016) 134 139

Materials were fed in a batch mode. The feed materials
were cut and cleaned in the cutting chamber through shear
and impact action, until the products passed the predefined
screen size apertures. The recycled GFRP consisted of fibrous
and powdered fractions. The products were retained in a
collection bin. For safety reasons, the transfer of materials
from the collection bin was performed while running an air
extractor to avoid dust inhalation by the operator.
3.2. Quality analyses of recyclates
Recyclates obtained in every experimental trial were
characterised through these following methods.
3.2.1. Fibre length measurement
The short fibres were randomly picked from each batch of
recyclates and were placed on a sticky tape before fibre length
distribution was characterised. Microscropic images of the
fibres were taken using a ProgRes C10 optical microscope.
Measurement of fibre length was carried out using ImageJ
software. For each batch, 200 short fibres were measured.
of the granulator. This is considered as the non-cutting or no
load power. After the panels were inserted, the sudden power
increments were attributed to action between the panels and
the granulator cutting blades. The increments were also
caused by additional power needed to reduce the initial panel
size into smaller fragments. The next phase is the minor
cutting where smaller fragments were recirculated and
cleaned until they can pass through the screen holes. During
this phase, power demand profile is dominated by the basic
power of the granulator. The processing of smaller fragments
does not require higher power demand compared to
processing of the initial panel size. The power needed for
cutting operation (tip power) was determined by finding
difference between average value of the basic power and the
power demand recorded during the cutting phases. The overall
energy demand is the area under the power versus processing
time graph.

3.2.2. Sieving

Sieving is required to classify the recyclates into different

size grades. In this study, the sieving process was performed
using W.S. TYLER mechanical sieve shakers (RX-812-3)
with mesh sizes of 5.00 mm, 2.36 mm, 1.40 mm, 0.43 mm
and 0.15 mm. The shaker was set for 10 minutes of sieving
time. The machine vibrated the column of sieves horizontally
to allow materials smaller than the mesh size to fall into the Fig. 2. Typical power demand profile throughout mechanical recycling
lower sieve stage. The weight of recyclates retained on each process using a Wittmann MAS1 granulator
sieve stage was recorded. It should be noted that only
favourably aligned fibers can be successfully separated with 4.2. Specific tip and total energy demand
this method. Vibration and prolonged sieving time helps
alleviate this challenge. From environmental perspective, the aim of this study is to
determine optimum granulator parameters to minimise overall
3.2.3. Furnace treatment energy demand per kilogram of material processed. Figure 3
and 4 show overall specific tip and total energy demand for
Residual resin content was determined by burning off the each experimental run. The energy demand was normalised to
organic portion of the recyclates in a furnace. For each unit mass of feed material since each experimental run has a
experimental trial, finer recyclates retained at the sieve size of wide range of sample weight as a result of various panel size
0.43 mm diameter were put in a ceramic crucible. The and thickness.
crucible had a small open lid to allow evolved gases to
escape. Based on initial trials, the furnace temperature was set
to 400 C for 2 hours followed by 600 C for 30 minutes. The
temperatures were chosen to ensure the matrix part of the
recyclates was completely degraded. The difference in the
sample weight before and after the furnace treatment
represents percentage of the residual resin content.

4. Results and discussions

4.1. Power demand profile of Wittmann MAS1 granulator

A typical power profile of MAS1 granulator in mechanical

recycling of composites is illustrated in Figure 2. The power Fig. 3. Specific tip energy demand for all experiments
demand before panel insertion was taken as the basic power
 Norshah Aizat Shuaib and Paul Tarisai Mativenga / Procedia CIRP 48 (2016) 134 139 137

Comparing Figure 3 and 4, specific tip energy only depend on desired recyclate size which is closely related to
accounted of less than 1% of the specific total energy demand. the associated reuse applications.
The finding is in agreement with machining studies in
literature, where power demand from machine auxiliary units 4.3. Quality of recyclates
and basic functions dominate the overall energy usage [23,
24]. From Figure 3, experiment T3 and T4 have high specific The recyclates from the granulator trials were in forms of
energy demand. Bigger plate thickness used in both flakes, short fibres and powder fractions. These include
experiments was the main reason. It is reasonable to expect materials retained on the screen after the process and the ones
more cutting energy is required for the thick panel which retained in the collection bin. Fibre length distribution was
means high un-deformed chip thickness. determined to predict the reinforcement effect in new
As can be seen from Figure 4, experiment T3 has the composite products. The results are illustrated in Figure 5.
greatest energy demand per kilogram, followed by experiment Most of the fibres have length less than 1.50 mm. The initial
T1. These two experiments used smaller screen size of 4 mm. fibre length was 50 mm in the original material.
The small screen size prolonged the processing time required
to reduce size of recyclates until the material can pass through
the screen holes. As energy is a time dependent quantity, long
processing time leads to an increment in total specific energy
demand. Greater value for experiment T3 is also attributable
to combination of large plate size and thickness.

Fig. 5. Fibre length distribution in recyclates

Classification of size for the recyclates was carried out

based on weight fraction. As a result from the sieving trials,
the materials were further classified into coarse (recyclates
retained on 2.36 mm sieve stage and above) and fine grades
Fig. 4. Total specific energy demand for all experiments (recyclates below the 2.36 mm sieve stage), as shown in
Figure 6. The coarse grade predominantly consisted of large
Main effect plots for mean specific tip and total energy flakes. The coarse fractions may be considered too large for
demand were generated using Minitab 16 statistical software immediate reuse applications and should be re-processed. The
using minimum the better objective. P-value for each factor fine fractions were the useful portion of the recyclates for the
and response were determined using ANOVA analysis based reuse applications.
on 95% confidence interval. The p-values and interpretation
are shown in Table 2. Factors with p-value less than 0.05 are
considered to be statistically significant and have major
contribution to the energy performance of the granulator. The
result shows that all process parameters have statistically no
significant influence on the specific tip energy demand. The
only strong contribution is the screen size for specific total
energy demand, with a p-value of 0.003. This can be related
to influence of screen size on the overall processing time
which directly influence the total energy usage. Fig. 6. Physical appearance of recyclates after a sieving trial
Table 2. Summary of ANOVA result (p-value) for energy demand for each (a) Coarse; (b) Fine
parameter based on 95% confidence level (=0.05).
P-value and result interpretation
The weight percentages from the sieving trials are shown
Process
parameters
in Table 3 and Figure 7. It is apparent that in all experiments,
Specific tip energy Specific total energy
the coarse fractions accounted for around 50-70% of the
Plate thickness 0.079 (NS) 0.745 (NS) weight percentages. Within the fine fractions, the highest
Screen size 0.419 (NS) 0.003 (S)
Plate size
weight percentage is between 0.43 mm and 1.40 mm sieve
0.346 (NS) 0.220 (NS)
size. Experiment T3 trials have low weight percentage of the
S: Significant, NS: Not significant
coarse fractions.
Clean recycled fibres with low residual resin content are
Taken together, these results suggest that choice of screen
important to ensure good interfacial bonding with matrix in
size is vital in optimising the specific energy demand in
new composite products. Residual resin contents were
mechanical recycling of composites. However, this has to
138 Norshah Aizat Shuaib and Paul Tarisai Mativenga / Procedia CIRP 48 (2016) 134 139

determined for the sieved recyclates between the size of 0.43 will allow shear and impact action during the cutting process
mm and 1.40 mm and the results are included in Figure 8. The to be performed efficiently, hence preventing the presence of
resin content in the recyclates from all experiments has large particle size in the output.
comparable values, with average resin content of 61%. The
original sample has about 70% of matrix by weight which Table 4. Summary of ANOVA result (p-value) for recyclate quality for each
parameter based on 95% confidence level (=0.05).
gives resin elimination efficiency of about 13%. The resin
elimination in mechanical recycling is considerably lower P-value and result interpretation
Process
compared to 100% using chemical recycling method [13]. parameters Mean fibre Weight fraction of Resin
length fine recyclates content
Table 3. Weight fraction of recyclates after sieving trials Plate thickness 0.564 (NS) 0.016 (S) 0.273 (NS)
Screen size 0.634 (NS) 0.199 (NS) 0.441 (NS)
Sieve size (mm) Average weight percentage of recyclates (%) Plate size 0.557 (NS) 0.204 (NS) 0.089 (NS)
T1 T2 T3 T4 S: Significant, NS: Not significant
> 5.00 61.16 57.82 27.25 47.86
2.36 < x < 5.00 12.86 16.28 26.63 20.95 4.4. Comparison between two granulator technologies
1.40 < x < 2.36 6.39 7.67 13.85 7.97
0.43 < x < 1.40 11.87 13.45 21.43 14.81
0.15 < x < 0.43 5.31 4.55 9.67 6.99 A further study was carried out to compare performance of
< 0.15 2.41 0.24 1.16 1.42 the Wittmann MAS1 granulator to the Eco-Wolf grinder
Model GM-2411-50 (Figure 9 a)). The Eco-Wolf grinder has
a capacity of 800 kg/hour with a rated motor power of 37 kW.
The clearance gap between the granulator blades and screen
was 3 mm. The cutting chamber of the grinder is shown in
Figure 9 b). Materials recovered from the Eco-Wolf grinder
were supplied by Eco-Wolf Inc. The recycling process used 5
mm screen size. The processed material was GFRP scrap
from a boat manufacturer with 5.5 mm thickness. The same
material was processed using the Wittmann MAS1 granulator
with the same screen size. Figure 10 shows the recyclates
recovered from each machine. The recyclates from the Eco-
Wolf grinder (lower clearance gap) has less amount of large
particle size.
Fig. 7. Weight percentage of coarse and fine fractions in recyclates

Fig. 9. Eco-Wolf grinder Model GM-2411-50 (a) machine; (b) rotor

blades and cutting chamber (courtesy of Eco-Wolf Inc.)

Fig. 8. Residual resin content in recyclates

The ANOVA analysis was used to determine the

significance of the results using 95% confidence interval. The
results are shown in Table 4. Only plate thickness was found
to be statistically significant in affecting the weight fraction of
coarse and fine grades in the recyclates. The results show that
there is no dominant effect of the operational parameters on
mean fibre length and resin content in the recyclates. The
significant influence of the plate thickness in the Wittmann
MAS1 granulator process can be attributed to the clearance
gap between the blades and the under screen. For the case of
recycling composite waste, the machine design requires feed
materials to have a thickness close to the clearance gap. This
Fig. 10. Physical appearance of recyclates (a) Eco-Wolf GM-2411-50
grinder; (b) Wittmann MAS1 granulator
Results for processes energy demand and recyclate quality
are included in Table 5. At the maximum processing capacity,
the specific energy demand for the Eco-Wolf grinder is less
due to greater processing rate of 800 kg/hour compared to
only 30 kg/hour for the Wittmann granulator. Recyclates from
 Norshah Aizat Shuaib and Paul Tarisai Mativenga / Procedia CIRP 48 (2016) 134 139
the Eco-Wolf process has low resin content, higher mean fibre
length and higher generation of processed or fine fractions.
These positive outcomes can be attributable to a low clearance
gap of 3 mm in Eco-Wolf grinder, compared to 5 mm
clearance gap in the Wittmann granulator. A low clearance
gap will allow more material to be processed therefore more
fibrous fraction can be produced. This also reduces particle
size and the generation of flakes. This finding provides an
important insight on the influence of machine design on
quality of composite recyclates.
Table 5. Comparison of performances between Wittmann MAS1 granulator
and Eco-Wolf GM-2411-50
Process output
Wittmann MAS1 Eco-Wolf GM-2411-50
Specific energy
0.32 MJ/kg 0.14 MJ/kg [25]
demand at maximum
(at 30 kg/hour) (at 800 kg/hour)
processing rate
Mean fibre length 0.82 mm 1.21 mm
Residual resin content 65.2% 47.7%
Weight fraction of Coarse (82.4%) Coarse (2.9%)
recyclates Fine (17.6%) Fine (97.1%)
5. Conclusions
This study investigated the effect of operational
parameters on energy consumption and associated recyclate
quality in mechanical recycling of GFRP. Main conclusions;
x The total energy demand of the granulator is dominated by
its basic energy. The direct specific total energy is highly
dependent on the granulator screen size. Small screen size
prolongs the material residence time in the cutting
chamber hence increases the energy demand. It is
important to capture impact of processing rate on energy
demand in recycling processes. This information can
improve the fidelity of life cycle datasets.
x Material thickness was found to have a significant role in
determining power requirements and how efficiently the
material is being cut. Thicker materials led to more
effective cutting and higher fraction of fine, cleaned or
useful recyclates.
x There is a clear trade-off between the process energy
demand and recyclate quality. Larger screen size can
reduce process energy demand but the recyclates will
contain high weight fraction of coarse particles which
requires re-processing steps. Sorting recyclates by fibre or
matrix size will enable particular applications.
x Granulator design, particularly the machine clearance gap
is an important factor in determining the recyclate quality.
The composite plate thickness should be close to the
machine clearance gap to ensure efficient cutting process.
Acknowledgements
The authors acknowledge the UK Engineering and Physical Sciences
Research Council (EPSRC), grant EP/K026348/1, Efficient X-sector use of
HeterogeneoUs MatErials in Manufacturing (EXHUME). We are grateful to
Production Glassfibre and Eco-Wolf Inc. in providing materials and
information for this study.
Data Access
This publication is accompanied by data from the experiments (see
http://dx.doi.org/10.15127/1.299002). Detailed procedure explaining how this
data was collected and analysed is provided in Section 2 and 3 of this paper.
139
References
[1] D. Gay, S.V. Hoa, Composite Materials: Design and Applications, CRC
Press 2003.
[2] Ernst & Young, UK Composites Supply Chain Scoping Study, in, 2010.
[3] E. Witten, The European GRP Market, in: Composites Market Report
2014: Market developments, trends, challenges and opportunities, 2014.
[4] A. Conroy, S. Halliwell, T. Reynolds, Composite recycling in the
construction industry, Composites Part A: Applied Science and
Manufacturing, 37 (2006) 1216-1222.
[5] M. Bains, E. Stokes, Developing a Resource Efficiency Action Plan for
the Composites Sector, in, URS & Netcomposites, 2013.
[6] Directive 2000/53/EC, Directive on End of Life Vehicles,, in, Official
Journal of the European Communities, 2000, pp. 34-55.
[7] G. Oliveux, L.O. Dandy, G.A. Leeke, Current status of recycling of fibre
reinforced polymers: Review of technologies, reuse and resulting properties,
Progress in Materials Science, 72 (2015) 61-99.
[8] S.J. Pickering, Recycling technologies for thermoset composite
materialscurrent status, Composites Part A: Applied Science and
Manufacturing, 37 (2006) 1206-1215.
[9] J. Palmer, O.R. Ghita, L. Savage, K.E. Evans, Successful closed-loop
recycling of thermoset composites, Composites Part A: Applied Science and
Manufacturing, 40 (2009) 490-498.
[10] A. Torres, I. De Marco, B.M. Caballero, M.F. Laresgoiti, M.J. Chomn,
G. Kondra, Recycling of the solid residue obtained from the pyrolysis of
fiberglass polyester sheet molding compound, Advances in Polymer
Technology, 28 (2009) 141-149.
[11] D. kesson, Z. Foltynowicz, J. Christen, M. Skrifvars, Microwave
pyrolysis as a method of recycling glass fibre from used blades of wind
turbines, Journal of Reinforced Plastics & Composites, 31 (2012) 1136-1142.
[12] S.J. Pickering, R.M. Kelly, J.R. Kennerley, C.D. Rudd, N.J. Fenwick, A
fluidised-bed process for the recovery of glass fibres from scrap thermoset
composites, Composites Science and Technology, 60 (2000) 509-523.
[13] G. Oliveux, J.-L. Bailleul, E.L.G.L. Salle, Chemical recycling of glass
fibre reinforced composites using subcritical water, Composites Part A:
Applied Science and Manufacturing, 43 (2012) 1809-1818.
[14] S. Feih, E. Boiocchi, G. Mathys, Z. Mathys, A.G. Gibson, A.P. Mouritz,
Mechanical properties of thermally-treated and recycled glass fibres,
Composites Part B: Engineering, 42 (2011) 350-358.
[15] D. Rouholamin, Y.T. Shyng, L. Savage, O.R. Ghita, A comparative
study into mechanical performance of glass fibre recovered through
mechanical grinding and high voltage pulse power fragmentation, in: 16 th
European Conference on Composite Materials, Seville, Spain, 2014.
[16] Hohenstein Institute, Biotechnology innovation helps with recycling
carbon fibres (press information), http://www.hohenstein.de/media /pdf/635
EN_06_Recycling_Carbonfasern_2015_101440.pdf. (accessed: 08/2015).
[17] S. Job, Recycling glass fibre reinforced composites history and
progress, Reinforced Plastics, 57 (2013) 19-23.
[18] A.C. Meira Castro, M.C.S. Ribeiro, J. Santos, J.P. Meixedo, F.J.G. Silva,
A. Fiza, M.L. Dinis, M.R. Alvim, Sustainable waste recycling solution for
the glass fibre reinforced polymer composite materials industry, Construction
and Building Materials, 45 (2013) 87-94.
[19] V.S.P. Bitra, A.R. Womac, N. Chevanan, P.I. Miu, C. Igathinathane, S.
Sokhansanj, D.R. Smith, Direct mechanical energy measures of hammer mill
comminution of switchgrass, wheat straw, and corn stover and analysis of
their particle size distributions, Powder Technology, 193 (2009) 32-45.
[20] S. Mani, L.G. Tabil, S. Sokhansanj, Grinding performance and physical
properties of wheat and barley straws, corn stover and switchgrass, Biomass
and Bioenergy, 27 (2004) 339-352.
[21] Z. Miao, T.E. Grift, A.C. Hansen, K.C. Ting, Energy requirement for
comminution of biomass in relation to particle physical properties, Industrial
Crops and Products, 33 (2011) 504-513.
[22] Wittmann Battenfeld, Models MAS1 to MAS4. http://www.wittmann-
group.com.au/wp-content/uploads/2014/07/WITTMANN-Granulator-MAS-
1-to-MAS-4.pdf [accessed: 04/2015].
[23] K. Salonitis, P. Ball, Energy Efficient Manufacturing from Machine
Tools to Manufacturing Systems, Procedia CIRP, 7 (2013) 634-639.
[24] V.A. Balogun, P.T. Mativenga, Modelling of direct energy requirements
in mechanical machining processes, Journal of Cleaner Production, 41 (2013)
179-186.
[25] Eco-Wolf Inc., Energy consumption details, Factory data, No date.